Figure 1: 3D Scatter Plot: Robbery Count vs. Socio-Economic Indicators

Plotly

The relationship between Permanent_Job_and_Labour_Force_Ratio, Individual_Median_Average_Income and robbery count among neighbouhoods is shown in Figure 1: Toronto 158 Neighbourhoods’ Robbery Count vs. Socioeconomic Indicators. There is a noticeable cluster at low individual median income, and surprisingly with a high permanent job labour force ratio.

Figure 2: Boxplots with Facets: Robbery Count By Premises Type and Darkness

Plotly

The distribution of robbery counts by Premises_Type, categorized based on the presence of sunset, Darkness is shown in Figure 2: Robbery Counts by Premises Type and Darkness. We can see that the mean and variance of robbery count in darkness are clearly higher than those not in darkness, regardless of premises type. Additionally, the darkness group has more outliers and a higher robbery count than the other group. Looking through premises types, we can see that commercial and educational areas have much bigger ranges in robbery count for the 1st quartile and 3rd quantile in the darkness group. This may imply that people who appear in those areas may have a higher probability of being involved in a robbery crime during darkness compared to the other group.

Figure 3: Line Plots with Facets: Robbery Count by Premises Time and Temporal Factors

Plotly

The changes in robbery counts in different Premises_Type across different Season, Month, Day_of_Week, and Hour of the day is shown in Figure 3. We can see that commercial and outdoor areas have higher counts than all other types of premises.

Regarding the season, the count of robbery increases sharply in winter, which is consistent with existing studies that show crime rates increase during national holidays like Christmas and New Year.

For the month, the count of robbery shows an increasing trend during the summer months, which also matches the conclusion of existing research suggesting a correlation between temperature and crime rates. Regarding the day of the week, we can see that robbery counts increase during weekends and drop on weekdays.

For hours of the day, it is clear that the robbery count increases from the afternoon to midnight and decreases sharply at 2 AM, which aligns perfectly with my intuition in creating the Darkness variable.

Season

Month

Day of Week

Hour of Day

Figure 4: Barplots with Facets: Total Crime Counts and Robbery Proportion by Temporal Factors

Plotly

The total crime counts and proportion of robbery crime across different seasons, months, days of the week, and hours of the day are shown in Figure 5. We can observe that not only do the total crime counts increase during holidays, weekends, and darkness, but also the proportion of robberies.

Season

Month

Day of Week

Hour of Day

Figure 5: Interactive Map: Map of Toronto Robbery Crimes by Premises Type in 2020

Leaflet

The geographical distribution of robbery incidents across Toronto is shown in Figure 5. Each marker on the map represents a robbery incident, with color indicating the type of premises where the incident occurred. Popup information provided details such as neighborhood name, premises type, and population.

We can observe clusters of robbery crimes on the map, as well as areas that do not have any reported robberies. This suggests that there may be a spatial influence on the probability and occurrence of robbery crimes.